Field Comparison of Tensiometer and Granular Matrix Sensor Automatic Drip Irrigation on Tomato

Abstract

A low-volume/high frequency (LVHF) soil moisture-based drip irrigation system was tested on a shallow sandy soil at a commercial tomato (Lycopersicon esculentum) farm in southern Florida. Six LVHF irrigation treatments were compared with the standard commercial practice on the farm (control), where a portable pump was used for manual drip irrigation twice each week. In the six LVHF treatments the system was continuously pressurized by means of an electrical pump and a pressure tank, and controlled by an irrigation timer set to irrigate a maximum of five times per day with the irrigation time (i.e., volume) set according to historical evapotranspiration (ET) demands in the area. Two treatments were based on timer schedules, one to supply 100% of the maximum recommended crop water needs in the area based on historical ET (ET-100%), and the other to supply 150% of those needs (ET-150%). The other four treatments were created by interfacing two types of soil moisture sensors (switching tensiometers and granular matrix sensors with control modules) set at two moisture points (wet = 10 kPa, optimal = 15 kPa) in a closed control loop with the irrigation timer programmed at the ET-100% schedule. Results showed that the six LVHF treatments reduced water use while not significantly affecting tomato yields. Switching tensiometers at the 15 kPa set point performed the best (up to 73% reduction in water use when compared to the control, 50% with respect to ET-100%). The results show that water use below historical ET levels can be obtained without sacrificing yield by keeping the root zone moisture at controlled levels with the soil-moisture based system. Routine maintenance was critical for reliable operation of the switching tensiometers. Granular matrix sensor based irrigation behaved erratically, and did not improve water savings compared to ET-100%, indicating that this system was not effective under the conditions of the area due to the sensor's slow response to frequent wetting-rewetting cycles and characteristics of the interface.